Take Rate: How to Set It, Taper It and Defend It
Once both sides know each other, what stops them dealing directly? If the answer is nothing, you have an introduction service.
Summary
- Charge the side that gains most from the match and has the fewest alternatives, and judge the rate by contribution per transaction, not the headline.
- Taper the rate as seller volume rises, keep the tiers few and public, and add a floor fee for small transactions. A flat rate pushes your best sellers toward their own channel.
- A rate charged purely for access is the most vulnerable there is. Justify it with ongoing services (payments, trust, demand, financing) so that leaving costs a seller something every time.
Your is the single most consequential number in a marketplace. It decides your revenue, your margin, how fast each side grows, and whether either side eventually tries to go around you.
And it usually gets set the way most pricing gets set: by looking at what a comparable platform charges and picking something similar.
There are four questions underneath it, and they deserve better answers than that.
Question one: who pays?
Supply, demand, or both.
This is not only an economics decision. It changes who feels like your customer.
Charge supply and your sellers experience you as a cost of doing business, while buyers experience you as free. This is the common structure, and it's usually right where supply has more to gain and demand is harder to acquire.
Charge demand and buyers feel the price, which raises the bar on what you must deliver to them. It works where buyers get something they clearly can't get elsewhere.
Charge both and you have two pricing conversations and two sets of expectations. Legitimate in categories where both sides genuinely benefit, and harder to run.
The general rule: charge the side that gains most from the match and has the fewest alternatives. But check it against who you find harder to acquire, because that side needs the gentler treatment.
Question two: how much?
Enough to build a business. Low enough that neither side starts looking for a way around you.
There's no universal number. Rates vary enormously by category depending on what the platform actually does. A marketplace that only connects two parties cannot charge what one that handles logistics, payments, quality assurance and disputes can.
What matters more than the headline rate is your contribution per transaction. Take rate minus payment costs, support, fraud, logistics if you carry it, and whatever incentives you're paying out. Most marketplaces we look at can quote their take rate instantly and have to go and calculate the second number. It is the same floor every other business prices against.
If contribution per transaction is negative, growth makes you worse off, and no amount of volume fixes it.
Question three: does it move?
A flat rate regardless of volume is the easiest to explain and the worst structure you can choose.
Here's why. Your largest sellers generate the most value for you, cost you the least per transaction to serve, and have the most to gain by leaving. A flat rate charges them exactly what it charges a seller doing a tenth of the volume, which quietly pushes your best supply toward building their own channel.
A tapering rate, lower as a seller's volume rises, does three things at once. It rewards your best supply. It gives sellers a reason to consolidate volume onto you rather than spreading it. And it makes leaving expensive, because they'd give up a rate they've earned.
Keep the tiers few and the thresholds public. A tapering structure that only some sellers know about becomes a negotiation, and then you're back to a flat rate with extra steps.
A minimum fee is the other half of this. Below a certain transaction value, a percentage doesn't cover your cost to process and support it. A floor fee protects the economics of small transactions without distorting the headline rate. Choosing what you charge per is the same decision in a different business.
Question four: what justifies it?
This is the question that decides whether your take rate survives, and it's the one most founders think about last.
A rate charged purely for access is the most vulnerable rate there is. Access is a one-time benefit. Once the buyer and seller know each other, the introduction has already happened, and every subsequent transaction on your platform is one they could have done directly.
A rate justified by ongoing services is different, because leaving costs the seller something real every time.
What counts:
- Payments and settlement. You carry the risk, you handle the reconciliation, and for many sellers a reliable settlement cycle is worth more than a lower rate.
- Logistics and fulfilment, where you provide it.
- Quality assurance and trust. Verification, ratings, dispute resolution, guarantees. This is what makes a buyer willing to transact with a seller they don't know, and it is the single most defensible thing a marketplace does.
- Demand generation. If you're spending to bring buyers that a seller couldn't reach alone, that's a service with a measurable value.
- Financing. Working capital, advances against receivables, credit to buyers. Increasingly where Indian marketplaces find defensible margin.
- Tools and data. Inventory management, analytics, pricing insight, catalogue support.
The test is simple: if a seller left tomorrow, what would they have to rebuild? If the honest answer is "nothing much, they'd just message their buyers directly," your rate is at risk regardless of how reasonable it looks.
The disintermediation problem, honestly
Every marketplace lives with this, and most founders reach for the wrong answer.
The wrong answer is enforcement: contact details hidden, penalties for off-platform transactions, terms of service that forbid it. These slow it down and irritate the people you were trying to keep. They don't solve it, because you cannot police a phone call.
The right answer is making the platform genuinely more useful than the direct relationship. Payment protection the direct deal doesn't have. Dispute resolution. A record they need. Financing. Scale of demand. Convenience that outweighs the saving.
Every rupee of your take rate should be answerable with "and here's what that buys."
Published rate versus realised rate
A point that catches most marketplaces.
Say your published take rate is 8%. Now subtract seller incentives, buyer coupons, category promotions, waived fees for strategic sellers, payment gateway costs and the settlement cycle you're funding.
Your realised rate is meaningfully lower, and almost nobody tracks the gap.
Build the waterfall, published rate down to what actually reaches you per transaction, the same way a manufacturer would build one from list price to net realisation. It's the same exercise and it produces the same surprise. On the operating side, the match rate and repeat rate tell you whether the rate is buying what it should.
Subsidies need an exit before they start
Almost every marketplace subsidises one or both sides early. That's reasonable.
What isn't reasonable is doing it for years without defining when it ends.
Three questions to answer before you start a subsidy, not after:
- What is it buying? Liquidity in a specific city, supply in a specific category, a habit among a specific group of buyers. "Growth" is not an answer.
- What does success look like? The measurable state at which it's no longer needed: a match rate, a repeat rate, a density of supply at which the market works on its own.
- When do we taper, and by how much? With a date.
Without those three, a subsidy stops being an investment and becomes a permanent cost everyone has learned to call growth.
What else can a marketplace charge for?
Your take rate doesn't have to be your only revenue line, and in mature marketplaces it usually isn't.
- Advertising and placement. Sellers paying for visibility. High margin, and it can quietly become the real business.
- Subscriptions. A monthly fee for enhanced tools, better placement or a lower commission rate. Gives you predictable revenue and gives sellers a reason to commit.
- Value-added services priced separately: logistics, photography, cataloguing, financing, insurance.
- Listing or setup fees, which also filter out sellers who were never serious.
The reason to diversify isn't only revenue. Each additional service you provide is another thing a seller would have to rebuild if they left, which makes your core take rate more defensible, not less.
Changing a take rate
Harder than changing a price, because both sides notice at once.
- Grandfather existing sellers with a stated end date. Open-ended creates two classes of seller and a structure nobody can explain.
- Change it alongside something you're adding, not on its own. A rate change paired with a new service is a different conversation from a rate change on its own.
- Move your smallest sellers first if you can. They have the least bargaining power and the least to lose, and you learn what happens before you approach the ones who matter most.
- Expect your largest sellers to negotiate. Have a position ready, and know your contribution per transaction for each of them before the conversation starts.
Where to start on your own take rate
Pick your three largest sellers and your three largest categories. Work out your contribution per transaction for each: realised rate, after every cost and incentive.
Then ask, for each one: if they left tomorrow, what would they have to rebuild?
Those two answers together tell you whether your take rate is defensible or simply unchallenged so far.
Our Pricing Maturity Assessment covers structure and defensibility alongside your cost floor.
Frequently asked questions
What is a good marketplace take rate?
It depends entirely on what the platform does. A marketplace that only makes introductions cannot charge what one handling payments, logistics, quality assurance and disputes can. The more useful number is contribution per transaction: take rate minus payment costs, support, fraud, logistics and incentives.
Should a take rate be flat or tiered?
Tiered, tapering as seller volume rises. A flat rate charges your most valuable sellers the same as your smallest, and those are exactly the sellers most able to build their own channel. A taper rewards them and makes leaving expensive.
How do I stop sellers and buyers going around the platform?
Not through enforcement: you can't police a phone call. Make every transaction on the platform genuinely better than the direct one: payment protection, dispute resolution, financing, records, demand they couldn't reach alone. Every part of your take rate should be answerable with what it buys.
How useful was this article?
One tap. It tells us what to write more of.
The next one
Get what we publish next, by email.
Working notes on raising, borrowing, protecting, growing and structuring capital in India. One email a week at most, and you can leave any time.
We use your address only to send this. See our privacy policy.
We store your address to send you these emails and nothing else. See our privacy policy.
Related reading
